Go Beyond Explanatory Analysis With Your Geospatial Data

In order to stay competitive, companies are turning to more advanced forms of analysis, like location analytics, to answer their important business questions. While feature-rich mapping tools allow us to visualize things like economic downturns, misaligned territories, and poorly constructed shopping lanes, using the tool without any thought of how the data is constructed leads to nothing more than an expensive visualization tool.

So, how do you make the most of your location analytics tool?  It starts with data prep.

Geocoding:

When prepping data, the first problem to tackle is often geocoding. Geocoding is the process of turning addresses into latitude and longitudes. This is a daunting task for many, but the good news is we have a few tips to get you started.

The first question you need to answer is what kind of geocoding you need. Do you need to plot a customer’s point or ‘rooftop’ address, or is knowing the city they live in sufficient?

• If city is sufficient, search the web for ‘free city lat/lng’. You will find several large data sets that provide names and the geographic center of cities. Using your BI tool, associate the geographic center of each city to the city name in the address field of your data. Now, all customers with an address that includes the name of a city will be linked to a single point in the center of the city. This allows you to show the number of addresses represented in the single geographic point using size or color.
• If you need a rooftop address, you’ll need a larger data set to query. Companies like Google, MapQuest, and Willowbend offer large databases that match address to lat/lng pairs around the world. Often these are free up to a certain number of conversions per month. You might need a larger set of monthly conversions, but we often see customers paying for more conversions than they need. Since most rooftop geocoding providers charge per call to their database, make sure to only call new distinct addresses, which will drastically decrease the cost of goecoding. For example, MapQuest offers 15,000 free calls per month, which means you would have to attain 15,000 new customers each month before getting charged for geocodes.

If you are exceeding the free tier, we can help assess your current processes and recommend more cost-effective solutions.

Building Custom Territories:

After the geocoding problem is tackled, the next issue is building custom territories. Finding state or postcode polygons online is pretty simple since most countries provide them for free on a government site. While helpful, many companies have custom territories that group similar business areas together, so these free polygons don’t meet their needs. Before seeking a new tool to create custom territories, we advise our customers to first check their current tools. Many don’t realize that all major database systems – MS SQL Server, PostGres, Oracle, MySQL, etc. – have the functionality to create custom territories with the ability to programmatically modify the territories anytime there is a realignment.

TAKE YOUR GEOSPATIAL ANALYSIS FARTHER

We often see customers stop working on their geospatial data set after geocoding and custom territories are completed, but this is where they should start their analysis. Geocoding shows where existing customers are, and custom territories show an already defined sales strategy. Now, using some geospatial techniques, easily accessible prospects can be uncovered and territories can be scrutinized.

Here are some scenarios in which we helped our customers use their geospatial data to drive their business in new ways:

Point-in-Polygon to Optimize Delivery and Pick-Up Routes

Scenario: One of our customers offers office waste shredding by deploying trucks to client sites for waste pick-up. Each time they gained a new customer, they went through a labor-intensive onboarding process to create a new route that accommodated the new customer while trying not to disrupt the current routes as much as possible.

To ease this route-building process for new customers, we automated the process of locating the closest route with the option to see a 5-, 10- and 15-minute drive time perimeter around each route. Now, our customer can programatically identify if a new client falls within an acceptable drive time perimeter of an existing route. If they do, the client is automatically assigned to a specific route, saving our customer time and limiting the overhead of expensive detours. Furthermore, using the same drive time perimeters, our customer was able to execute a marketing campaign targeting new prospects along existing routes, allowing for a lower cost customer gain.

“Nearest Neighbor” to Create Custom Territories

Scenario: Using postcodes to create custom territories (for sales, distribution, etc.) and simply grouping customers within nearby postcodes together doesn’t always meet the needs of some companies. Using a technique called “nearest neighbor,” companies can instead dynamically group their customers together based on other criteria.

We worked with a company that has many store locations in very localized markets. They could not group postcodes to create sales territories surrounding their stores, because they did business in many non-contiguous postcodes.

Store Locations (Blue Dots) Grouped by Postcode

Creating sales territories based on postcodes would misrepresent the distribution of their customer base. Distributing their customer base within postcode territories meant some customers would travel unncessarily far when there was a closer store in a different postcode.

If Grouped By Postcode, The Customer (Red Dot) Would Be Assigned To A Store Location (Blue Dot) Unnecessarily Far Away

To solve this, we drew voronoi polygons around each of their store locations. A voronoi polygon is a set of the smallest set of contiguous polygons possible where all boundaries of the polygon are equadistant from the next point (in this example, the store location/blue dot). Thus, the voronoi polygon represents the smallest possible localized market based on store location.

Using Voronoi Polygons, Customer Territories Are More Logically Distributed. The Customers (Red Dots) Are Assigned To Store Locations (Blue Dots) That Are Close By.

This company can now determine which of their customers fall within the voronoi polygon and establish localized markets that are more logically distributed.

There are many ways to aggregate voronoi polygons – one simple approach is to group them based on the customer count in an area so that you can see how customers are distributed across an area.

Make the most of your geospatial data

These are just a few of the tricks we’ve learned to make the most of geospatial data. Visit QlikMaps.com to access videos and guides about QlikMaps..

Questions about QlikMaps? Contact Us